Kumu / Corporate Finance / Stage 2
Stage 2 — Company Selection Memo
docs/decisions/
Choose the company you'll live with all semester and defend the choice the way a senior analyst writes to a managing director: concise, decision-oriented, evidence-tight.
⚠ Do this first
Add adamwstauffer as a Write collaborator on your repo:
Settings → Collaborators → Add people → choose Write. Without it the instructor
can't leave tracked PR feedback. Sixty seconds, do it now.
What you'll learn
- How to frame an analytical project's scope before touching data
- What makes a hypothesis falsifiable — and why "we'll see what the ratios show" earns nothing
- How professional review works: PR-based feedback, the same loop as a manager marking up a draft memo
Picking a company that won't fight you
Eligibility rules exist to keep the analysis clean:
- Publicly traded on a recognized exchange — HOSE, HNX, SGX, SET, IDX, Bursa, PSE, NYSE, NASDAQ all qualify. Vietnamese and ASEAN listings are encouraged.
- An observable share price. You value the company in Stages 6 to 8 and set your value per share against its market price, so the shares must trade and you must be able to cite a dated closing price (the exchange, the company's investor-relations page, or a market-data site). An unlisted subsidiary, a delisted firm, or a stock suspended from trading does not qualify.
- Audited annual report available with enough line-item detail to populate the template.
- Non-financial company. Banks, insurers, and REITs have structurally different balance sheets — inventory turnover means nothing to a bank.
- At least 2 years of comparable data — remember the
startYear_*ranges need the prior year.
Pro move
Before committing, actually open the annual report and check you can find: total revenue, COGS, interest expense, inventory, receivables, total debt, and shareholders' equity — for both years. If any are missing or buried, pick a different company now, not at Stage 3.
The six required sections
| # | Section | What "strong" looks like |
|---|---|---|
| 1 | Company Overview | Name, ticker, exchange, industry, market cap, reporting currency — in three sentences |
| 2 | Selection Rationale | Why this company is analytically interesting: recent M&A, capital-structure shift, industry disruption, relevance to your career |
| 3 | Data Availability & Sources | Named sources ("HOSE disclosure portal", "company IR page"), fiscal year end, reporting standard (VAS / IFRS / U.S. GAAP) |
| 4 | Preliminary Observations | 2–3 hypotheses in "I expect X because Y" form — directional and falsifiable |
| 5 | Ratio Categories Preview | Which categories matter most for this industry and why |
| 6 | Data Collection Plan | Statements needed, market assumptions to source, currency/standard considerations |
Anatomy of a hypothesis
✗ "We will see what the leverage ratios show." ← not falsifiable
✓ "I expect leverage ratios to rise from FY2024 to
FY2025 because the company financed its regional
expansion with USD-denominated debt issued in
early 2025." ← direction + reasonA good hypothesis can be wrong. That's the point — at Stage 5 you'll check each one against the computed ratios, and a cleanly falsified hypothesis makes a better final analysis than a vague one that can't lose.
Using an LLM — reviewer first, then collaborator
Draft the memo yourself in the template first — your company, your reasons, your hypotheses. Then have an LLM review it, and only after that work it over together. The division of labor:
| LLM does well, once your draft exists | You must do |
|---|---|
| Checking all six sections against the template's structure | Choose the company — that's your judgment |
| Stress-testing your "I expect X because Y" hypotheses | Verify eligibility (listed, non-financial, 2+ years of data) |
| Identifying likely data sources | Confirm those sources actually have the data |
| Tightening toward senior-analyst tone | Edit for accuracy — the LLM doesn't know your company like you should |
Log any prompt beyond a two-line nudge in prompt-log.md — the log is
graded at Stage 4, and starting the habit now is free practice.
Submission checklist
adamwstaufferadded as Write collaborator (−5/stage until done)- Memo at
docs/decisions/YYYY-MM-DD-{company-slug}-selection-memo.md— all lowercase - YAML frontmatter from the template kept intact
- All six required sections present
- Hypotheses written in "I expect X because Y" form
- Verified the annual report actually contains the needed line items, both years
- AI sessions logged in
prompt-log.md
Rubric
The headings below are what the work is read for. Weights and due dates are in your course schedule.
| Criterion |
|---|
| Company selection & rationale |
| Analytical framing & hypotheses (falsifiable, directional) |
| Data source identification |
| Professionalism & communication |
Plus: Stage 5 grades how you incorporated the instructor's PR feedback on this memo. Expect comments; plan to respond visibly — accept, modify, or push back with reasons. Silence reads as oversight, not judgment.